Weighting, limited bandwidth Pink PN, room gain?

[thewire]

I setup my levels using 75dB limited bandwidth pink noise with my disc, then I went into REW and am checking levels using Pink PN from the CD player. I needed to raise my subs gain from -9 to -7, about where my other speakers are at. After my filters I raise the gain 1dB back to where it was relative to the level of my mains that the REW SPL meter is telling me. When I check in the RTA with Pink PN I burned using REW at -20dB FS I am looking for what I expected to be a flat looking line. I used my SPL Meter calc file. I'm supposed to use that right? Why does it look like I have a house curve when I calibrate the sub relative to the mains? I thought when looking at an RTA it is suppossed to be looking like a flat response in our graphs if we calibrate our subs flat, position our mains for a flat response. It would look like this if I were to measure a sweep also after setting levels. Please explain to someone that is new at this.

[Wayne A. Pflughaupt]

Not sure I understand the question, but if this is what your in-room response looks like with a regular REW sweep, it's going to look the same with the RTA. The only difference is that the former is a "snapsnot" reading, while the latter is real time.

Regards,
Wayne

[thewire]

Wayne A. Pflughaupt wrote:

Not sure I understand the question, but if this is what your in-room response looks like with a regular REW sweep, it's going to look the same with the RTA. The only difference is that the former is a "snapsnot" reading, while the latter is real time.

Regards,
Wayne

Yes that is it what it looked like last time I did a sweep. Prior to moving all my subs to the walls midpoints, the Pink PN would give me a flat looking response after I calibrated the mains and the subs. I would measure with the sweep and make some minor adjustments at the crossover then but it didn't look like this except silightly similar if I were to measure in the back of my room. Last time I checked with my limited bandwidth Pink Noise after calibrating with Pink PN, the levels were perfect, I had 75dB for the mains, and 74dB for the subs. This was after equalization with a -9 setting on my pre/pro subwoofer output. This was when I had a sub in each corner instead.

Now when I calibrate 75dB with the same limited bandwidth pink noise, I get a higher looking response for the subwoofers, and the VU meter is at -18 out -18 in instead of the previous -18 out, -24 in after calibrating the setup in REW.

I don't understand why if I calibrate the sub flat, then add the mains, the mains look lower, and the sub now also looks higher. Isn't the sub going to look flat with the sub+mains target if it tracks the subwoofer target no matter which gain or sub position is used, same with adding the mains?

[thewire]

I reversed the polarity on all my subs and the REW target is now tracking more similar to the measurements I am getting using Pink Noise. There is also the gain at 80Hz - 90Hz from my the length of my room which is just making things way to confussing. I don't think that I will be able to fix this problem with room treatments but that sure would be nice. I may end up getting a DCX2496 and try some custom crossover slopes. It's that or both eq my mains and subs at the same areas. The closer I sit to the center of the room, the greater that it is a problem.

[thewire]

Why does an 80Hz crossover look like this in both my DVD player, and the Outlaw 950's multichannel inputs? Flipping the switch on or off on the pre/pro does nothing.

When I use a crossover from the CD input it looks like a 12dB octave crossover that will correctly track the mains target, if it were 12dB octave, and the crossover for the sub will look like a correct sub target tracking the 24dB octave crossover, but the level is so low by the time I turn it up to match the mains crossover is overlapping? This all seems to be causing some cancellation around 100Hz where these crossovers overlap.

When I engage a crossover of 60Hz in the CD input and output to the subs then measure that, without using eq they track a 80Hz crossover at 24dB octave slope. By that time so much is taken from the mains that there is a huge hole between the two.

Any ideas? A 120Hz or 150Hz crossover in the player and the 80Hz in the pre/pro looked pretty good. Is the DCX2496 going to correct the crossovers so they look normal? Please excuse the wavy lines I used my power adapter for the laptop.

6-channel analog input
bass management off: full-range speakers receive full-range signal, subwoofer receives LFE signal plus information below 80Hz from the other five full-range speakers ("double bass" condition); subwoofer also has a filter that discards material above 120Hz on the LFE input signal (material that subs typically can't reproduce well)
bass management on: full-range speakers receive all information above 80Hz, subwoofer receives LFE signal plus information below 80Hz from other five full-range speakers; subwoofer also has a filter that discards material above 120Hz on the LFE input signal (material that subs typically can't reproduce well)

Great. What does this mean?

[brucek]

Great. What does this mean?

What part don't you understand?
It standard bass management.

brucek

[Wayne A. Pflughaupt]

thewire wrote:

This all seems to be causing some cancellation around 100Hz where these crossovers overlap.

When I engage a crossover of 60Hz in the CD input and output to the subs then measure that, without using eq they track a 80Hz crossover at 24dB octave slope. By that time so much is taken from the mains that there is a huge hole between the two.

It's unclear if you're seeing this by taking readings from the electronics or in-room from the speakers, but what you're describing is perfectly normal behavior for electronic crossovers. Depending on the alignment (Bessel, Butterworth, Linkwitz-Riley) and slope, you can see drops in response at the crossover frequency. Linkwitz-Riley filters have become the defacto standard in modern electronics, because their characteristics deliver flat response at the crossover frequency. So generally monkeying around with mismatched crossover frequencies is unnecessary. That said...

There is also the gain at 80Hz - 90Hz from my the length of my room which is just making things way to confussing. I don't think that I will be able to fix this problem with room treatments but that sure would be nice. I may end up getting a DCX2496 and try some custom crossover slopes.

...occasionally there will be a situation like this, where under-lapping the crossover frequencies can eliminate a room problem that occurs in the crossover range.

Regards,
Wayne

[thewire]

Sorry, I was not aware standard bass management meant tracking a crossover higher in REW than I would set. Yes these are the measurements directly from the pre/pro and I get similar results from the player. My reponse from the subwoofers tracks the same target which looks like a 120Hz crossover.

What I don't understand -

- I need to track a 120Hz crossover in REW and apply filters to this. The subwoofer should track the target in the pre/pro. Then what does this mean for my crossover without sending LFE information? I guess there is no way to check this then. What I will be looking at is a response of LFE + mains. There is no way around this although it is not common to see LFE in the 100Hz - 120Hz area. I need the subs to be 120Hz 24dB octave, the mains to be 100Hz cossover 12dB octave.

- Bass management does not send a full range signal to the mains in REW when I turn this switch off. Something seems wrong if that is what it is suppossed to do. I'm not sure what they mean by "double bass". I have never measured a difference eather way, but perhaps I had overlooked something.

[thewire]

If I were to do something about the crossover area, which if I'm understanding looks correct, this is what I have then to work on, or not. This is without adjusting the level of the sub. When I adjust the levels of the mains, the sub level changes. I calibrated them using the REW SPL meter.

[thewire]

For reason my 6 channel bypass and multichannel inputs are toast.

[thewire]

To get reference level I must now turn my subwoofer down from -7 to -10 using the CD inputs instead of the multichannel inputs and turn down the level of the pre/pro from 0 to -10. I had measured a similar looking response using the optical cable. Here are the impulse responses from some measurements using the mulitchannel or more recently the CD input. If this can somehow help to determine what may have went wrong, please let me know.

Measurement of speakers and sub using CD input no eq



Measurement of speakers and sub using no eq with bypass



Measurement of speakers and sub using eq with bypass

[Wayne A. Pflughaupt]

Okay, I'm totally lost. I can't weed out the relevant info from the TMI in these posts. Most people get satisfactory results by turning off their mains, EQing their sub(s) using REW, then blending the sub level with the mains, and it's a done deal. Not sure why that's not working here...

Regards,
Wayne

[thewire]

Sorry allow me to explain this in steps.

- First my subs were out-of-phase. Being so a result in REW tracked my 75 target when I would adjust the mains gain relative to the subs gain. That's a no brainer but can't be determined by me without switching the polarity on my subs and mains, then checking to see if eather the mains is in phase by using a voltmeter, or if I somehow notice things are different. A good indication should have been having to turn up the gain on the sub level, but I overlooked it. The short and simple. When all speakers are out-of-phase, measurements will look different. Instead of room gain at the area where pink-noise is located, your going to get room cancellation, and the gain must go up.

- The mulitchannel inputs don't work anymore. I'm trying to determine why. In no way was I suggesting anything to do with eq versus not eq in this thread.

- I was earlier trying to determine why my mulitichannel inputs with bypass did not work as they were advertised sending a full range signal to the mains, and why the standard 80Hz crossover did not match REW 80Hz 24dB octave which was what I read the Outlaw 950 featured. It also did not match the same 80Hz crossover in the CD input.

- Decided to just try the best I could, probobly just causing some confussion in the proccess but I thought things were running more smoothly, until my inputs failed to work any longer.

I'm now using the CD input instead, which matches levels relative to my optical connection.

[thewire]

The multichannel bypass is now working again, but I have my DVD player currently set to bitstream instead of PCM, which if I understand correctly enables High Def audio formats to be internally decoded in the player if I set that to PCM. The switch on the back is set to the off postion for the bass management, but it was not working for this setting earlier as well.

[thewire]

Here Wayne are some measurements of the CD versus the mulitchannel input I am having trouble with. My pre/pro was at -10 sub gain at -10. Now that I measure with a frequency response instead of directly from the pre/pro I notice dramatic changes at the crossover area.

Green = CD input
Blue = Multichannel bypass input with switch for bass management off
Purple = Mulitichannel bypass input with switch for bass managment on

Example one with a main speakers + or - wires connected



Example two reversing the - or + wires on the main



edit: They mains are in phase in the second image. The Levels are matching on the subs. The problem I see is a lack of diffussion caused by the symmetry in my room. How this would apply to the front of my room and what I could do (for home theater) about it, I am unsure so if equalization of the 300Hz - 400Hz area will help, I'm willing to take that plunge. I don't see much reason to eq the subwoofers but may give it a try. To get bypass to work again I changed all my analog input setups to digital instead.

[thewire]

The dip at 160Hz was caused by over-absorption at critical points in key critical areas of phase related to the ports of the speakers near the floor. There was also some cancellation caused by this at the wall nearest to the speakers, with reflections off the floor from this travelling as high as 2/3 the way up the walls.

I placed reflection behind the speaker against the stage that was plywood (under other bass traps) using concrete with 2" panels over it, adjusted absorption panels at first reflection points, and the dip went up fast. The level of the mains also measure higher from my DVD player so that is factoring in what level looks like after I use a DVD to set levels. They actually match and do not need adjustment compared to when looking at the level lower in measurements, which I had ended up trying to correct. It is easier to turn down the subwoofer level, which I have done instead for this graph. This I was able to do using some eq, and bass management on. Dependent on other settings or proccessing not related to the bypass, I somehow had different results. I'm leaving it at that for that subject.